Last night I gave my talk at a meeting on financial regulation. The meeting was organized by David Evans. I gather that his goal is to put together a group of practitioners in the field of finance and regulation to influence public policy in the wake of the crisis. If I had been in his position, I am not sure I would have invited me to speak, but I’m glad I had the opportunity. Some observations:

[UPDATE: Tim Geithner was not at the meeting. His testimony is much less nuanced than the discussion at our meeting. He favors the sort of “neat” solutions that I believe would make the problem worse. He shows no sign of understanding the way in which his proposals could create new moral hazards (as “too big to fail” gets hard-wired into the system) and new points of vulnerability (by trying to make the system harder to break, you make it harder to fix).]There were about 35 people in attendance. Several were mid-to-upper-level officials at regulatory agencies (not agency heads, by any means). Several worked for law firms. Several worked for trade associations. My sense is that there were twice as many who were primarily lawyers rather than economists, although everyone seemed conversant with economics.

Given the demographics, one would have expected my free-market position to be pretty much ignored. Instead, I would describe the positions at the meeting as falling on a scale as follows:

1. Fixers. These are people who believe in regulatory reform, including creating a “systemic-risk regulator.” Their view is that our regulatory system has gaps. Nobody was responsible for “connecting the dots” between subprime excess and systemic fragility. Nobody was responsible for figuring out what AIG was doing to the financial system as a whole. We need to figure out what went wrong and fix it.

2. Cynics. These are people who are skeptical that regulatory reform will be a panacea. However, they take the view that Congress is bound to do something, and we might as well try to point them in the direction of doing something that does more good than harm.

3. Stubborn free-marketers. These are people who think that regulatory reform probably will do more harm than good, and therefore we oppose it.

If you had asked me to predict where the discussion would go ahead of time, I would have said that it would have stayed mostly between (1) and (2). Perhaps it would have been more productive had it done so. Instead, there was a lot of discussion between (2) and (3), which in some ways for me was a pleasant surprise. Why did this happen?

–selection bias. Evans is not a stubborn free-marketer, but his network within the legal/regulatory community is more market-oriented than is normal within that community.

–group dynamics. Even though I counted only two of us in the stubborn free-market camp, we were very passionate, while the “believers” were very cautious and ready to acknowledge problems and uncertainties with their proposed solutions. The result was that people who in their hearts were mostly between (1) and (2) felt inclined to make comments that fell between (2) and (3).

–the phenomenon of blaming the person not in the room. The person not in the room was Congress, so everyone felt comfortable bashing Congress and its response to political pressure. In retrospect, I wish I had raised some doubts about the notion that Congress is the problem (not that I am a fan of Congress). But if the financial-policy elite wants to do X and Congress is inclined to do Y, the elite is far from powerless. After all, look at TARP, which passed despite public pressure against it. It’s not as if Congress was eager to pass a bailout (although I think that the Democrats were excited about the leverage it would give the government over the financial sector).

My suggestion of trying to create a system that is easy to fix rather than one that is hard to break was sufficiently well understood that it generated some discussion. Afterward, one participant said that a good example would be the Soviet T-34 tank (easy to fix) compared with the German Tiger tank (a better machine when it worked, but often unrecoverable when it broke down). A waiter who was passing by during this conversation heartily agreed, and said he had used that same example a couple hours earlier.

I think easy-to-fix might include:

1. Returning to America’s traditional hostility toward large banks. Perhaps onerous regulation of interstate banking, in order to encourage more state banks.

2. Have market share limits on financial institutions that are for the purpose of reducing risk of a single failure. This is related to but different from anti-trust laws, whose purpose is to promote competition.

But above all, I think the key is to avoid enforcing “neat” solutions. From a hard-to-break perspective, it looks as though it would be easier to regulate a system where functions are neatly separated–mortgage lenders are not investment banks, insurance companies are not commercial banks, etc. From that perspective, you want a clean regulatory org chart, without any overlaps or duplication.

But from an easy-to-fix perspective, you don’t want financial institutions to be perfectly specialized. Instead, you want some messiness in the market, so that when one set of institutions fails, other institutions can step in and take their place. By the same token, having multiple regulators with overlapping responsibilities is a good thing, because you want institutional variety, not homogeneity.

Another after-dinner side conversation was about the future of economics. A few of us talked about focusing less on mathematical modeling and econometrics, with more focus on human behavior. In that context, I remarked that if you made me the systemic risk regulator, the one variable I would try to control would be the gender of the CEOs at “systemically important” financial institutions. I believe that changing that variable would do more to reduce risk than even the most ingenious design for regulatory architecture.